EXTENDED-RANGE FORECAST OF ATLANTIC SEASONAL HURRICANE
ACTIVITY AND LANDFALL STRIKE PROBABILITY FOR 2010
We foresee a very active hurricane season in 2010. We have increased our forecast from
early April, due to a combination of a transition from El Niño to currently-observed
neutral conditions and the continuation of unusually warm tropical Atlantic sea surface
temperatures. We anticipate a well above-average probability of United States and
Caribbean major hurricane landfall. All factors are lining up for a very active 2010
hurricane season.
(as of 2 June 2010)
By Philip J. Klotzbach 1 and William M. Gray 2
This forecast as well as past forecasts and verifications are available via the World Wide
Web at http://hurricane.atmos.colostate.edu/Forecasts
Emily Wilmsen, Colorado State University Media Representative, (970-491-6432) is
available to answer various questions about this forecast
Department of Atmospheric Science
Colorado State University
Fort Collins, CO 80523
Email: amie@atmos.colostate.edu
1
2
Research Scientist
Professor Emeritus of Atmospheric Science
1
Why issue extended-range forecasts for seasonal
hurricane activity?
We are frequently asked this question. Our answer is that it is possible to say
something about the probability of the coming year’s hurricane activity which is superior
to climatology. The Atlantic basin has the largest year-to-year variability of any of the
global tropical cyclone basins. People are very curious to know how active the upcoming
hurricane season is likely to be, particularly if you can show hindcast skill improvement
over climatology for many past years.
Everyone should realize that it is impossible to precisely predict this season’s
hurricane activity in early June. There is, however, much curiosity as to how global ocean
and atmosphere features are presently arranged as regards to the probability of an active
or inactive hurricane season based on similar conditions in past years. Our early June
statistical forecast methodology shows strong evidence over 58 past years that significant
improvement over climatology can be attained. We would never issue a seasonal
hurricane forecast unless we had a statistical model developed over a long hindcast
period which showed significant skill over climatology.
We issue these forecasts to satisfy the curiosity of the general public and to bring
attention to the hurricane problem. There is a general interest in knowing what the odds
are for an active or an inactive season. One must remember that our forecasts are based
on the premise that those global oceanic and atmospheric conditions which preceded
comparatively active or inactive hurricane seasons in the past provide meaningful
information about similar trends in future seasons. This is not always true for individual
seasons. It is also important that the reader appreciate that these seasonal forecasts are
based on statistical schemes which, owing to their intrinsically probabilistic nature, will
fail in some years. Moreover, these forecasts do not specifically predict where within the
Atlantic basin these storms will strike. The probability of landfall for any one location
along the coast is very low and reflects the fact that, in any one season, most U.S. coastal
areas will not feel the effects of a hurricane no matter how active the individual season is.
However, all coastal residents should prepare for an active hurricane season every year.
Landfalling tropical cyclones can devastate communities in inactive or active seasons. It
only takes one landfalling system to make this a very active season for you.
2
ATLANTIC BASIN SEASONAL HURRICANE FORECAST FOR 2010
Forecast Parameter and 1950-2000
Climatology (in parentheses)
Issue Date
9 December
2009
11-16
51-75
6-8
24-39
3-5
6-12
100-162
108-172
Named Storms (NS) (9.6)
Named Storm Days (NSD) (49.1)
Hurricanes (H) (5.9)
Hurricane Days (HD) (24.5)
Major Hurricanes (MH) (2.3)
Major Hurricane Days (MHD) (5.0)
Accumulated Cyclone Energy (ACE) (96.1)
Net Tropical Cyclone Activity (NTC) (100%)
Issue Date
7 April
2010
15
75
8
35
4
10
150
160
Issue Date
2 June
2010
18
90
10
40
5
13
185
195
PROBABILITIES FOR AT LEAST ONE MAJOR (CATEGORY 3-4-5) HURRICANE
LANDFALL ON EACH OF THE FOLLOWING UNITED STATES COASTAL
AREAS:
1) Entire U.S. coastline - 76% (average for last century is 52%)
2) U.S. East Coast Including Peninsula Florida - 51% (average for last century is
31%)
3) Gulf Coast from the Florida Panhandle westward to Brownsville - 50% (average
for last century is 30%)
PROBABILITY FOR AT LEAST ONE MAJOR (CATEGORY 3-4-5) HURRICANE
TRACKING INTO THE CARIBBEAN (10-20°N, 60-88°W)
1) 65% (average for last century is 42%)
3
ABSTRACT
Information obtained through May 2010 indicates that the 2010 Atlantic hurricane
season will be much more active than the average 1950-2000 season. We estimate that
2010 will have about 10 hurricanes (average is 5.9), 18 named storms (average is 9.6), 90
named storm days (average is 49.1), 40 hurricane days (average is 24.5), 5 major
(Category 3-4-5) hurricanes (average is 2.3) and 13 major hurricane days (average is 5.0).
The probability of U.S. major hurricane landfall and Caribbean major hurricane activity
is estimated to be well above its long-period average. We expect Atlantic basin Net
Tropical Cyclone (NTC) activity in 2010 to be approximately 195 percent of the longterm average. We have increased our seasonal forecast from early April.
This forecast is based on an extended-range early June statistical prediction
scheme that utilizes 58 years of past data. Analog predictors are also utilized. The
influence of El Niño conditions is implicit in these predictor fields, and therefore we do
not utilize a specific ENSO forecast as a predictor.
We expect that the current trend from El Niño to neutral conditions will persist
and that weak La Niña conditions will develop by the most active portion of this year’s
hurricane season (August-October). The expected trend towards weak La Niña
conditions should lead to reduced levels of vertical wind shear compared with what was
witnessed in 2009. Another reason for our forecast increase is due to the persistence of
anomalously warm sea surface temperatures in both the tropical and North Atlantic.
Current SST anomalies are running at near-record warm levels. These very warm waters
are associated with dynamic and thermodynamic factors that are very conducive for an
active Atlantic hurricane season. Another factor in our forecast increase is the weakerthan-normal Azores High that prevailed during April-May. Weaker high pressure
typically results in weaker trade winds that are commonly associated with more active
hurricane seasons. Another important factor is that we are in the midst of a multi-decadal
era of more major hurricane activity. Major hurricanes cause 80-85 percent of
normalized hurricane damage.
We are also debuting a hurricane forecast for activity in the Caribbean Basin.
This forecast is based on a statistical prediction scheme that utilizes 60 years of past data.
This model is predicting a very active season for the Caribbean.
4
How will the Gulf oil spill impact a hurricane?
We do not anticipate that the oil spill will have any noticeable impact on tropical cyclone
intensity or frequency. The strong winds of a tropical storm or hurricane should
sufficiently mix the oil and water that there should be no noticeable alterations in broadscale evaporation and sensible and latent heat flux.
What impact will a hurricane have on the Gulf oil spill?
This depends on the storm’s track in relation to the oil spill. If the storm tracks to the
west of the oil, there is the potential that the counter-clockwise circulation of the
hurricane could drive some of the oil further towards the U.S. Gulf Coast. Alternatively,
a storm tracking to the east of the oil could push the oil further offshore. But, little is
understood about the interaction of tropical cyclones and oil.
5
Notice of Author Changes
By William Gray
The order of the authorship of these forecasts was reversed in 2006 from Gray and
Klotzbach to Klotzbach and Gray. After 22 years (1984-2005) of making these forecasts,
it was appropriate that I step back and have Phil Klotzbach assume the primary
responsibility for our project’s seasonal forecasts. Phil has been a member of my
research project for the last ten years and was second author on these forecasts from
2001-2005. I have greatly profited and enjoyed our close personal and working
relationship.
Phil is now devoting much more time to the improvement of these forecasts than I
am. I am now giving more of my efforts to the global warming issue and in synthesizing
my projects’ many years of hurricane and typhoon studies.
Phil Klotzbach is an outstanding young scientist with a superb academic record.
I have been amazed at how far he has come in his knowledge of hurricane prediction
since joining my project in 2000. I foresee an outstanding future for him in the hurricane
field. He is currently developing new seasonal and two-week forecast innovations that
are improving our forecasts. The success of the last two years of seasonal forecasts is an
example. Phil was awarded his Ph.D. degree in 2007. He is currently spending most of
his time working towards better understanding and improving these Atlantic basin
hurricane forecasts.
Acknowledgment
We are grateful to the National Science Foundation (NSF) for providing partial
support for the research necessary to make these forecasts. We also thank the
GeoGraphics Laboratory at Bridgewater State College (MA) for their assistance in
developing the United States Landfalling Hurricane Probability Webpage (available
online at http://www.e-transit.org/hurricane).
The second author gratefully acknowledges the valuable input to his CSU
seasonal forecast research project over many years by former project members and now
colleagues Chris Landsea, John Knaff and Eric Blake. We also thank Professors Paul
Mielke and Ken Berry of Colorado State University for much statistical analysis and
advice over many years. We also thank Bill Thorson for technical advice and assistance.
6
DEFINITIONS AND ACRONYMS
Accumulated Cyclone Energy (ACE) - A measure of a named storm’s potential for wind and storm surge destruction defined as the
sum of the square of a named storm’s maximum wind speed (in 104 knots2) for each 6-hour period of its existence. The 1950-2000
average value of this parameter is 96.
Atlantic Multi-Decadal Oscillation (AMO) – A mode of natural variability that occurs in the North Atlantic Ocean and evidencing
itself in fluctuations in both sea surface temperature and sea level pressure fields. The AMO is likely related to fluctuations in the
strength of the thermohaline circulation. Although several definitions of the AMO are currently used in the literature, we define the
AMO based on North Atlantic sea surface temperatures from 50-60°N, 10-50°W.
Atlantic Basin – The area including the entire North Atlantic Ocean, the Caribbean Sea, and the Gulf of Mexico.
El Niño – A 12-18 month period during which anomalously warm sea surface temperatures occur in the eastern half of the equatorial
Pacific. Moderate or strong El Niño events occur irregularly, about once every 3-7 years on average.
Hurricane (H) - A tropical cyclone with sustained low-level winds of 74 miles per hour (33 ms-1 or 64 knots) or greater.
Hurricane Day (HD) - A measure of hurricane activity, one unit of which occurs as four 6-hour periods during which a tropical
cyclone is observed or is estimated to have hurricane-force winds.
Madden Julian Oscillation (MJO) – A globally propagating mode of tropical atmospheric intra-seasonal variability. The wave tends to
propagate eastward at approximately 5 ms-1, circling the globe in approximately 40-50 days.
Main Development Region (MDR) – An area in the tropical Atlantic where a majority of major hurricanes form, defined as 10-20°N,
20-70°W.
Major Hurricane (MH) - A hurricane which reaches a sustained low-level wind of at least 111 mph (96 knots or 50 ms-1) at some point
in its lifetime. This constitutes a category 3 or higher on the Saffir/Simpson scale.
Major Hurricane Day (MHD) - Four 6-hour periods during which a hurricane has an intensity of Saffir/Simpson category 3 or higher.
Multivariate ENSO Index (MEI) – An index defining ENSO that takes into account tropical Pacific sea surface temperatures, sea level
pressures, zonal and meridional winds and cloudiness.
Named Storm (NS) - A hurricane, a tropical storm or a sub-tropical storm.
Named Storm Day (NSD) - As in HD but for four 6-hour periods during which a tropical or sub-tropical cyclone is observed (or is
estimated) to have attained tropical storm-force winds.
Net Tropical Cyclone (NTC) Activity –Average seasonal percentage mean of NS, NSD, H, HD, MH, MHD. Gives overall indication
of Atlantic basin seasonal hurricane activity. The 1950-2000 average value of this parameter is 100.
Saffir/Simpson Scale – A measurement scale ranging from 1 to 5 of hurricane wind and ocean surge intensity. One is a weak
hurricane; whereas, five is the most intense hurricane.
Southern Oscillation Index (SOI) – A normalized measure of the surface pressure difference between Tahiti and Darwin.
Sea Surface Temperature – SST
Sea Surface Temperature Anomaly – SSTA
Thermohaline Circulation (THC) – A large-scale circulation in the Atlantic Ocean that is driven by fluctuations in salinity and
temperature. When the THC is stronger than normal, the AMO tends to be in its warm (or positive) phase.
Tropical Cyclone (TC) - A large-scale circular flow occurring within the tropics and subtropics which has its strongest winds at low
levels; including hurricanes, tropical storms and other weaker rotating vortices.
Tropical North Atlantic (TNA) index – A measure of sea surface temperatures in the area from 5.5-23.5°N, 15-57.5°W.
Tropical Storm (TS) - A tropical cyclone with maximum sustained winds between 39 (18 ms-1 or 34 knots) and 73 (32 ms-1 or 63
knots) miles per hour.
Vertical Wind Shear – The difference in horizontal wind between 200 mb (approximately 40000 feet or 12 km) and 850 mb
(approximately 5000 feet or 1.6 km).
1 knot = 1.15 miles per hour = 0.515 meters per second
7
1
Introduction
This is the 27th year in which the CSU Tropical Meteorology Project has made
forecasts of the upcoming season’s Atlantic basin hurricane activity. Our research team
has shown that a sizable portion of the year-to-year variability of Atlantic tropical
cyclone (TC) activity can be hindcast with skill exceeding climatology. These forecasts
are based on a statistical methodology derived from 58 years of past data. Qualitative
adjustments are added to accommodate additional processes which may not be explicitly
represented by our statistical analyses. These evolving forecast techniques are based on a
variety of climate-related global and regional predictors previously shown to be related to
forthcoming seasonal Atlantic basin TC activity and landfall probability. We believe that
seasonal forecasts must be based on methods that show significant hindcast skill in
application to long periods of prior data. It is only through hindcast skill that one can
demonstrate that seasonal forecast skill is possible. This is a valid methodology provided
that the atmosphere continues to behave in the future as it has in the past.
The best predictors do not necessarily have the best individual correlations with
hurricane activity. The best forecast parameters are those that explain that portion of the
variance of seasonal hurricane activity that is not associated with the other forecast
variables. It is possible for an important hurricane forecast parameter to show little direct
relationship to a predictand by itself but to have an important influence when included
with a set of 2-3 other predictors.
A direct correlation of a forecast parameter may not be the best measure of the
importance of this predictor to the skill of a 2-3 parameter forecast model. This is the
nature of the seasonal or climate forecast problem where one is dealing with a very
complicated atmospheric-oceanic system that is highly non-linear. There is a maze of
changing physical linkages between the many variables. These linkages can undergo
unknown changes from weekly to decadal time scales. It is impossible to understand
how all these processes interact with each other. No one can completely understand the
full complexity of the atmosphere-ocean system. But, it is still possible to develop a
reliable statistical forecast scheme which incorporates a number of the climate system’s
non-linear interactions. Any seasonal or climate forecast scheme should show significant
hindcast skill before it is used in real-time forecasts.
2
June Forecast Methodology
We developed a new June forecast scheme which was used for the first time in
2008. This scheme worked out quite well in predicting a very active season in 2008. It
predicted approximately average activity in 2009 (an over-forecast). Complete details on
the earlier June forecast schemes used from 1995-2007 are available in our June 2008
forecast (Klotzbach and Gray 2008).
2.1
Current June Statistical Forecast Scheme
8
We have found that using two spring predictors and our early April hindcast, we
can obtain early June hindcasts that show considerable skill over the period from 19502007. This new forecast model also provided a very accurate prediction for the 2008
hurricane season.
This new scheme was created by evaluating the two spring predictors using leastsquared regression along with our early April NTC forecast. The adjusted June NTC
forecast was calculated using the following equation.
Adjusted June NTC = 0.4 * (Preliminary June NTC Hindcast) + 0.6 * (Final April
NTC Hindcast).
In rapid changeover years from El Niño to La Niña or from La Niña to El Niño,
our early April statistical forecast will likely either over-estimate or under-estimate
activity. However, by early June of transition ENSO years, the atmosphere/ocean system
has often begun its transition to the new ENSO phase, and therefore our April/May
predictors for our early June forecast should provide us with more reliable information
about the activity likely to occur for the upcoming season than our early April forecast.
By definition, least-squared regression tends to be conservative, and therefore,
predicting large outlier events can be quite challenging. In order to help adjust for this
challenge, the hindcasts from the linear regression model were adjusted to the final
hindcast in the following manner.
The standardized value of each hindcast was calculated. These hindcasts were
then adjusted to the final hindcast by multiplying by the standard deviation of the
observations. Since the standard deviation of the observations is considerably larger than
the standard deviation of the hindcasts, this aids in the forecast of outlying events. Any
hindcasts that resulted in an NTC forecast of less than 30 NTC units were adjusted to 30
NTC units, since no season has had an NTC of lower than 30 units since 1950.
As mentioned before, our new statistical scheme shows enhanced levels of
hindcast skill, explaining 61 percent of the variance from 1950-2007 and 83 percent of
the variance from 1995-2007. We believe that we have solid physical links between
these predictors and the upcoming Atlantic basin hurricane season.
Table 1 displays our early June hindcasts for 1950-2007 using the new statistical
scheme, while Figure 1 displays observations versus NTC hindcasts. Our early June
hindcasts have correctly predicted above- or below-average seasons in 42 out of 58
hindcast years (72%). These hindcasts have had a smaller error than climatology in 31
out of 58 years (53%). Our average hindcast error is 31 NTC units, compared with 46
NTC units for climatology. This scheme also shows considerable stability when broken
in half, explaining 54 percent of the variance from 1950-1978 and 70 percent of the
variance from 1979-2007. This new scheme is also well-tuned to the multi-decadal
active hurricane periods from 1950-1969 and 1995-2007 versus the inactive hurricane
9
period from 1970-1994 (Table 2). Note how small the errors are for multi-decadal
prediction.
The forecast scheme has shown excellent skill from 1980-2009. The average
forecast error using a climatological forecast of 100 NTC units over this period was 52
units, while the average hindcast error using our June statistical forecast was only 27
units (about half as much).
Figure 2 displays the locations of the two 1 June (using April-May data)
predictors used in this scheme in map form. Please refer to Figure 2 of our early April
forecast for locations of predictors used in our early April prediction scheme. Table 3
lists the three (two new spring predictors and our early April prediction) predictors that
are utilized for this year’s June forecast. A more extensive discussion of current
conditions in the Pacific and Atlantic basins is provided in Sections 6 and 7, respectively.
10
Table 1: Observed versus hindcast NTC for 1950-2007 using the current June scheme.
Average errors for hindcast NTC and climatological NTC predictions are given without
respect to sign. Red bold-faced years in the “Hindcast NTC” column (2) are years that
we did not go the right way, while red bold-faced years in the “Hindcast improvement
over Climatology” column (5) are years that we did not beat climatology. The hindcast
went the right way with regards to an above- or below-average season in 42 out of 58
years (72%), while hindcast improvement over climatology occurred in 31 out of 58
years (53%). The real-time forecasts for 2008 and 2009 are also listed.
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Average
(1)
Observed NTC
247
148
103
127
127
207
68
86
144
96
93
230
41
116
184
86
140
102
47
182
64
91
35
53
83
92
86
47
82
97
130
113
38
31
80
106
37
46
117
130
100
58
67
52
35
222
192
54
169
182
134
135
83
175
232
279
85
99
112
(2)
Hindcast NTC
212
114
177
186
126
181
123
119
166
108
139
214
126
75
171
149
136
99
53
145
86
52
30
101
30
71
33
38
30
30
89
95
35
34
131
126
30
103
96
159
116
30
34
102
53
209
176
59
194
197
181
124
86
143
173
204
137
100
113
(3)
Observed minus
Hindcast
35
34
-74
-59
1
26
-55
-33
-21
-12
-46
16
-85
41
13
-63
4
3
-6
37
-21
40
5
-49
53
20
53
9
52
67
42
18
2
-3
-51
-20
7
-58
21
-30
-15
28
33
-50
-17
13
16
-5
-25
-16
-47
12
-2
32
59
76
-52
-2
31
(4)
Observed minus
Climatology
147
48
3
27
27
107
-32
-14
44
-4
-7
130
-59
16
84
-14
40
2
-53
82
-36
-9
-65
-47
-17
-8
-14
-53
-18
-3
30
13
-62
-69
-20
6
-63
-54
17
30
0
-42
-33
-48
-65
122
92
-46
69
82
34
35
-17
75
132
179
-15
-1
46
(5)
Hindcast improvement over
Climatology
112
14
-71
-33
26
81
-23
-19
23
-8
-39
114
-26
-25
71
-49
36
-1
47
45
14
-31
59
-1
-36
-12
-39
43
-35
-63
-11
-5
60
66
-31
-14
56
-3
-4
0
-15
15
0
-2
47
109
76
41
44
66
-14
24
14
43
73
104
-37
0
+15
2008
2009
162
69
181
108
-19
-39
62
-31
+41
-8
11
Figure 1: Observed versus hindcast values of NTC for 1950-2007.
Table 2: Hindcast versus observed average NTC for active vs. inactive multi-decadal
periods.
Years
1950-1969
(Active)
1970-1994
(Inactive)
1995-2007
(Active)
Average
Hindcast
NTC
Average
Observed
NTC
141
129
69
75
153
157
12
Figure 2: Location of spring predictors for our June extended-range statistical prediction
for the 2010 hurricane season.
Table 3: Listing of 1 June 2010 predictors using the June statistical model for the 2010
hurricane season. A plus (+) means that positive values of the parameter indicate
increased hurricane activity during the following year.
Predictor
1) Subtropical Atlantic Index (+): April-May SST (20-50°N, 15-30°W) (+) & May
SLP (10-35°N, 10-40°W) (-)
2) April-May 200 MB U (5-25°S, 50-90°E) (-)
3) Early April Hindcast (+)
2010 Forecast Values
+1.6 SD
+0.9 SD
102 NTC
There is also extended-range forecast skill from 1 June for United States hurricane
landfall probabilities. In the 15 out of 58 years where our current hindcast scheme
forecast NTC values that were above 150, we had more than twice as many hurricane (41
versus 20) landfalls and more than three times as many major hurricane (17 versus 5)
landfalls along the U.S. coastline when compared with the 15 out of 58 years where our
hindcast scheme gave NTC values that were below 60. For the Florida Peninsula and the
U.S. East Coast, the ratio between NTC hindcast values greater than 150 and below 60
were 25 to 9 for hurricanes and 9 to 1 for major hurricanes – a remarkable difference.
13
2.2
Physical Associations among Predictors Listed in Table 3
The locations and brief descriptions of the two spring predictors for our early June
statistical forecast are now discussed. It should be noted that both forecast parameters
correlate significantly with seasonal physical features that are known to be favorable for
elevated levels of hurricane activity. These factors are primarily related to AugustOctober vertical wind shear in the Atlantic Main Development Region (MDR) from 1020°N, 20-70°W as shown in Figure 3.
Figure 3: Vertical wind profile typically associated with (a) inactive Atlantic basin
hurricane seasons and (b) active Atlantic basin hurricane seasons. Note that (b) has
reduced levels of vertical wind shear.
For each of these predictors, we display a four-panel figure showing linear correlations
between values of each predictor and August-October values of sea surface temperature,
sea level pressure, 200 mb zonal wind, and 925 mb zonal wind, respectively. In general,
higher values of SSTA, lower values of SLPA, anomalous westerlies at 925 mb and
anomalous easterlies at 200 mb are associated with active Atlantic basin hurricane
seasons.
For more information about the predictors utilized in our early April statistical forecast
(used as 60% of our early June forecast), please refer to our early April 2010 forecast:
14
http://tropical.atmos.colostate.edu/Forecasts/2010/april2010/apr2010.pdf
1. Subtropical Atlantic Index (+): April-May SST (20-50°N, 15-30°W) (+) &
May SLP (10-35°N, 10-40°W) (-)
A combination of above-normal sea surface temperatures (SSTs) in the eastern
subtropical Atlantic and lower-than-normal sea level pressures in the subtropical Atlantic
are associated with a weakened Azores high and reduced trade wind strength during the
late spring (Knaff 1997). This combined index in April-May is strongly correlated with
weaker trade winds, lower-than-normal sea level pressures and above-normal SSTs in the
tropical Atlantic during the following August-October period (Figure 4). All three of
these August-October features are commonly associated with active Atlantic basin
hurricane seasons, through reductions in vertical wind shear, increased vertical instability
and increased surface latent and sensible heat fluxes, respectively. Stronger-than-normal
values of this index correlate quite well (~0.5) with active Atlantic basin tropical cyclone
seasons.
Predictor 2. April-May 200 MB U in the South Indian Ocean (-)
(5-25°S, 50-90°E)
Upper-level easterly wind anomalies in the South Indian Ocean during April-May are
associated with anomalously low sea level pressure and anomalous upper-level
divergence in the western tropical Pacific and anomalously high sea level pressure and
anomalous upper-level convergence in the eastern tropical Pacific. These features are
associated with an active Walker Circulation, which is typically observed in cool ENSO
years. Figure 5 displays the significant correlations that are achieved between values of
this predictor in April-May and August-October sea surface temperatures, sea level
pressure and 925 and 200 mb zonal wind anomalies, respectively. Note the anomalous
easterly winds that are typically observed at upper levels over the tropical Atlantic and
Caribbean in August-October when upper-level easterly wind anomalies exist in the
South Indian Ocean in April-May. These anomalous easterlies combined with anomalous
westerlies at 925 mb in the tropical Atlantic, reduce vertical wind shear across the
tropical Atlantic providing a more favorable environment for tropical cyclone formation
and intensification. Predictor values have been trending slightly more positive in this
region since the 1950s. We have removed the trend in zonal wind anomalies from our
predictor calculations to avoid a potentially non-physical lowering of forecast values, as
there is some uncertainty as to the quality of the NCEP/NCAR reanalysis data for upperlevel winds in the 1950s.
This predictor is located in an area that has not been considered in our previous early
June forecasts. Since this predictor is new, we have gone through extensive testing to
make sure that the predictor is valid. The predictor shows considerable stability when
evaluated over both the 1950-1989 period and the 1990-2007 period. It correlates with
NTC at -0.57 over the period from 1950-1989 and correlates with NTC at -0.60 over the
period from 1990-2007. The correlation with NTC over the full time period from 19502007 is -0.59 (Figure 6).
15
When we examined the top 10 years when the predictor had its highest values and
compared them with the bottom 10 years when the predictor had its lowest values,
considerable differences were evident. In the 10 years when the zonal winds had their
largest easterly anomalies, an average of 150 NTC units were observed, compared with
an average of only 66 NTC units in the 10 years when the zonal winds had their largest
westerly anomalies (Figure 7).
16
Figure 4: Linear correlations between the subtropical Atlantic index (Predictor 1) and
August-October sea surface temperature (panel a), August-October sea level pressure
(panel b), August-October 925 mb zonal wind (panel c) and August-October 200 mb
zonal wind (panel d). All four of these parameter deviations are known to be favorable
for enhanced hurricane activity.
17
Figure 5: Linear correlations between April-May 200 mb U in the South Indian Ocean
(Predictor 2) and August-October sea surface temperature (panel a), August-October sea
level pressure (panel b), August-October 925 mb zonal wind (panel c) and AugustOctober 200 mb zonal wind (panel d). All four of these parameter deviations are known
to be favorable for enhanced hurricane activity. Predictor values have been multiplied by
-1 to allow for easy comparison with Figure 4.
18
Figure 6: April-May values of Predictor 2 (pink line) and Atlantic basin NTC activity
(blue line). Note the strong negative correlation between the two curves.
Figure 7: Top 10/bottom 10 NTC ratios for Predictor 2. Note that much larger values of
NTC were observed when Predictor 2 had anomalous easterly winds.
19
3
Forecast Uncertainty
One of the questions that we are asked regarding our seasonal hurricane
predictions is the degree of uncertainty that is involved. Obviously, our predictions are
our best estimate, but there is with all forecasts an uncertainty as to how well they will
verify.
Table 4 provides our early June forecasts, with error bars (based on one standard
deviation of absolute errors) as calculated from hindcasts over the 1990-2007 period,
using equations developed over the 1950-1989 period. We typically expect to see 2/3 of
our forecasts verify within one standard deviation of observed values, with 95% of
forecasts verifying within two standard deviations of observed values. Note that even the
low value of one standard deviation below our real-time forecast is higher than the
climatological average for all forecast parameters.
Table 4: Model hindcast error and our 2010 hurricane forecast. Uncertainty ranges are
given in one standard deviation (SD) increments. Climatological average values from
1950-2000 are in parentheses in the right-hand column.
Parameter
Named Storms (NS)
Named Storm Days
(NSD)
Hurricanes (H)
Hurricane Days (HD)
Major Hurricanes
(MH)
Major Hurricane Days
(MHD)
Accumulated Cyclone
Energy (ACE)
Net Tropical Cyclone
(NTC) Activity
4
Hindcast
Error (SD)
3.8
18.3
2010
Forecast
18
90
Uncertainty Range – 1 SD
(67% of Forecasts Likely in this Range)
14.2 – 21.8 (9.6)
71.7 – 108.3 (49.1)
2.1
9.0
1.2
10
40
5
7.9 – 12.1 (5.9)
31.0 – 49.0 (24.5)
3.8 – 6.2 (2.3)
4.5
13
8.5 – 17.5 (5.0)
39
185
146 – 224 (96.1)
37
195
158 – 232 (100)
Caribbean Forecast Methodology
We have developed a new forecast for the Caribbean that we are debuting with
this early June prediction. Predictors for the Caribbean may be different than those used
for the entire Atlantic basin. We intend to explore additional region-specific forecasts in
the future, such as forecasts for the Gulf of Mexico or the higher latitude Atlantic.
We define the Caribbean to extend from 10-20°N, 60-88°W. This model attempts
to predict seasonal levels of Accumulated Cyclone Energy (ACE) generated in the
Caribbean. Through a combination of three predictors discussed in detail below, we can
issue a forecast that shows significant levels of hindcast skill. As is done with the overall
basinwide forecast, this Caribbean forecast will also be updated in early August.
20
4.1
Caribbean Statistical Forecast Scheme
We have found that using three spring predictors, we can obtain early June
hindcasts that show considerable skill over the sixty-year development period from 19492008.
This new scheme was created by evaluating three spring predictors using leastsquared regression. By definition, least-squared regression tends to be conservative, and
therefore, predicting large outlier events can be quite challenging. In order to help adjust
for this challenge, the hindcasts from the linear regression model were adjusted to the
final hindcast in the following manner:
The standardized value of each hindcast was calculated. These hindcasts were
then adjusted to the final hindcast by multiplying by the standard deviation of the
observations. Since the standard deviation of the observations is considerably larger than
the standard deviation of the hindcasts, this aids in the forecast of outlying events. Any
hindcasts that resulted in a negative ACE prediction were assigned a final ACE value of
0.
Our statistical scheme shows significant hindcast skill, explaining 43% of the
variance over the 1949-2008 period. Table 5 displays our early June hindcasts for 19492008 using the new statistical scheme, while Figure 8 displays observations versus ACE
hindcasts. Our early June hindcasts have correctly predicted above- or below-average
seasons in 47 out of 60 hindcast years (78%). These hindcasts have had a smaller error
than climatology in 42 out of 60 years (70%). Our average hindcast error is 8.4 ACE
units, compared with 12.5 ACE units for climatology. This scheme also shows
considerable stability when broken in half, correlating at 0.60 from 1949-1978 and 0.71
from 1979-2008. This new scheme is also well-tuned to the multi-decadal active
hurricane periods from 1950-1969 and 1995-2007 versus the inactive hurricane period
from 1970-1994 (Table 6). As with our entire Atlantic forecast, the forecast skill for the
Caribbean basin since 1980 has been particularly skillful. Forecast improvement over
climatology was notably high.
Figure 9 displays the locations of the early June predictors used in this scheme in
map form. Table 7 lists the three predictors that are utilized for this year’s June forecast.
The model predicted an ACE of 17 for the Caribbean in 2009, while the observed
value of ACE was 4 for the Caribbean in 2009.
21
Table 5: Observed versus hindcast Caribbean ACE for 1949-2008. Average errors for
hindcast ACE and climatological ACE predictions are given without respect to sign. Red
bold-faced years in the “Hindcast ACE” column (2) are years that we did not go the right
way, while red bold-faced years in the “Hindcast improvement over Climatology”
column (5) are years that we did not beat climatology. The hindcast went the right way
with regards to an above- or below-average season in 47 out of 60 years (78%), while
hindcast improvement over climatology occurred in 42 out of 60 years (70%).
Year
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
Average
(1)
Observed ACE
7
16
24
8
7
40
48
5
0
9
1
17
38
0
28
21
0
22
27
0
12
4
19
0
0
18
4
1
1
11
26
21
4
0
0
3
1
1
6
42
16
8
0
0
4
3
23
22
3
44
24
13
23
6
6
51
58
4
47
22
15
(2)
Hindcast ACE
8
29
23
23
4
20
47
11
2
21
13
26
35
41
33
23
3
21
15
8
18
12
0
0
0
3
0
0
21
14
12
17
15
0
0
3
0
3
9
23
19
22
0
0
0
0
23
17
6
23
37
29
27
8
23
34
39
30
25
40
16
(3)
Observed minus
Hindcast
0
-13
1
-15
3
20
1
-7
-2
-12
-12
-9
3
-41
-4
-2
-3
2
13
-8
-6
-8
19
0
0
15
4
1
-20
-2
14
3
-12
0
0
0
1
-2
-3
19
-2
-14
0
0
4
3
0
5
-4
21
-12
-16
-4
-2
-17
17
19
-25
22
-18
8.4
(4)
Observed minus
Climatology
-7
2
9
-7
-8
25
34
-10
-15
-5
-14
2
23
-15
14
6
-14
8
13
-14
-3
-11
5
-14
-14
4
-10
-14
-14
-3
11
6
-11
-15
-15
-11
-13
-14
-8
27
2
-7
-15
-15
-11
-12
9
8
-12
30
10
-1
9
-8
-9
36
43
-10
32
7
12.5
(5)
Hindcast improvement over
Climatology
7
-11
8
-8
5
5
33
3
12
-7
1
-7
20
-26
9
4
11
6
0
6
-4
3
-15
14
14
-11
6
13
-6
1
-3
3
-1
15
15
11
12
12
5
8
0
-7
14
15
7
9
8
3
8
8
-3
-15
5
6
-9
20
25
-15
11
-11
4.1
2009
4
17
-13
-11
-2
22
Figure 8: Observed versus hindcast values of Caribbean Basin ACE for 1949-2008.
Table 6: Hindcast versus observed average ACE for active vs. inactive multi-decadal
periods. Percentage differences from the climatological average (1949-2008) are in
parentheses.
Years
1949-1969
(Active)
1970-1994
(Inactive)
1995-2008
(Active)
Average Hindcast ACE
Average Observed ACE
20 (133%)
16 (107%)
7 (47%)
8 (53%)
26 (173%)
25 (167%)
23
Figure 9: Location of spring predictors for our Caribbean statistical prediction for the
2010 hurricane season.
Table 7: Listing of 1 June 2010 Caribbean Basin predictors using the June statistical
model for the 2010 hurricane season. A plus (+) means that positive values of the
parameter indicate increased hurricane activity during the following year.
Predictor
1) May SST (15-35°N, 15-30°W) (+)
2) May 925 MB U (7.5°S-2.5°N, 160°E-160°W) (-)
3) April-May 200 MB V (5-25°N, 60-85°W) (+)
2010 Forecast
Values
+2.1 SD
-1.8 SD
+1.6 SD
The Caribbean model is calling for a very active hurricane season in 2010. Our forecast
model is calling for an ACE of 58, which would be the highest recorded since 2005.
4.2
Physical Associations among Predictors Listed in Table 7
The locations and brief descriptions of the three spring predictors for our June Caribbean
statistical forecast are now discussed. It should be noted that all three forecast parameters
correlate significantly with seasonal physical features that are known to be favorable for
24
elevated levels of hurricane activity. Table 8 and 9 display correlations between each
predictor and August-October-averaged sea surface temperature, sea level pressure, 200
mb zonal wind and 925 mb zonal wind in the Main Development Region (MDR) and in
the Caribbean, respectively. Since many storms that generate large values of ACE in the
Caribbean form in the MDR, one would expect that these predictors would correlate with
physical features in both regions. Correlations that are significant at the 95% level using
a two-tailed Student’s t-test are highlighted in bold-faced type.
Table 8: Correlations between 1 June Caribbean predictors and August-October values of
Main Development Region (10-20°N, 20-70°W) sea surface temperature, sea level
pressure, 200 mb zonal wind and 925 mb zonal wind.
Predictor
1) May SST (15-35°N, 15-30°W) (+)
2) May 925 MB U (7.5°S-2.5°N, 160°E-160°W) (-)
3) April-May 200 MB V (5-25°N, 60-85°W) (+)
SST
0.66
-0.05
0.30
SLP
-0.40
0.37
-0.37
200 MB U
-0.36
0.40
-0.40
925 MB U
0.37
-0.19
0.47
Table 9: Correlations between 1 June Caribbean predictors and August-October values of
Caribbean (10-20°N, 60-88°W) sea surface temperature, sea level pressure, 200 mb zonal
wind and 925 mb zonal wind.
Predictor
1) May SST (15-35°N, 15-30°W) (+)
2) May 925 MB U (7.5°S-2.5°N, 160°E-160°W) (-)
3) April-May 200 MB V (5-25°N, 60-85°W) (+)
SST
0.65
-0.03
0.25
SLP
-0.51
0.30
-0.25
200 MB U
-0.49
0.39
-0.23
925 MB U
0.42
-0.49
0.34
As was done with our Atlantic basin forecast, for each of these predictors, we display a
four-panel figure showing linear correlations between values of each predictor and
August-October values of sea surface temperature, sea level pressure, 200 mb zonal
wind, and 925 mb zonal wind, respectively. In general, higher values of SSTA, lower
values of SLPA, anomalous westerlies at 925 mb and anomalous easterlies at 200 mb are
associated with active Caribbean basin seasons.
1. May SST in the Subtropical Atlantic (+)
(15-35°N, 15-30°W)
Above-normal sea surface temperatures (SSTs) in the eastern subtropical Atlantic are
associated with a weakened Azores high and reduced trade wind strength during the late
spring (Knaff 1997). This predictor is very strongly positively correlated (0.65
correlation) with Caribbean basin sea surface temperatures during August-October
(Figure 10 and Table 9). Lower pressures are typically experienced throughout the
Caribbean as well, implying a very favorable thermodynamic environment for tropical
cyclone formation and intensification in this area during the climatologically most active
portion of the hurricane season. This predictor also correlates favorably with reduced
vertical wind shear in the Caribbean as well as above-average SSTs, below-average SLPs
and reduced vertical wind shear over the MDR.
25
Predictor 2. May 925 mb U in the Central Tropical Pacific (-)
(7.5°S-2.5°N, 160°E-160°W)
Enhanced trade wind strength in the central tropical Pacific is typically associated with a
La Niña event. Stronger-than-normal trades encourage mixing and upwelling, promoting
cooling. In addition, Kelvin wave propagation and associated warm water transport from
the western Pacific to the eastern Pacific is inhibited by this stronger easterly flow.
Consequently, strong correlations exist between values of this predictor and Nino 3.4
values. The correlation between this predictor and the May Nino 3.4 value is 0.60, the
correlation between this predictor and the June-July Nino 3.4 values is 0.64, and the
correlation between this predictor and the August-October Nino 3.4 value is 0.55. ENSO
is a critical player in Caribbean tropical cyclone activity, as alterations in vertical shear
caused by ENSO are more significant here than in any other part of the Atlantic basin
(Klotzbach 2010, manuscript submitted to J. Climate). There is a significant negative
correlation (-0.49) between this predictor and trade wind strength during August-October
and a significant positive correlation (0.39) between this predictor and 200 mb wind
strength during August-October over the Caribbean, indicating the considerable influence
that this predictor has on the vertical shear values experienced over the Caribbean (Figure
11 and Table 9).
Predictor 3. April-May 200 MB V in the Southwest Atlantic (+)
(5-25°N, 60-85°W)
Anomalous southerly flow in the southwest Atlantic Ocean is associated with an
anomalous upper-level trough and surface high pressure over South America,
consequently driving strong trades over the tropical eastern Pacific and promoting La
Niña conditions. Also, southerly flow at upper levels in the Caribbean is indicative of an
upper-level anticyclone to the east. This upper-level anticyclone imparts anomalous
easterly upper-level flow, weakening the climatological upper-level westerlies and
thereby weakening vertical shear over both the tropical Atlantic as well as the Caribbean.
It appears that this predictor’s impacts are more strongly related to conditions over the
MDR as opposed to the Caribbean, as its correlations are statistically significant with
SST, SLP, 200 MB U and 925 MB U over the MDR (Figure 12 and Table 8), while
displaying only a significant correlation with 925 MB U over the Caribbean (Figure 12
and Table 9).
26
Figure 10: Linear correlations between subtropical Atlantic SST (Predictor 1) and
August-October sea surface temperature (panel a), August-October sea level pressure
(panel b), August-October 925 mb zonal wind (panel c) and August-October 200 mb
zonal wind (panel d). All four of these parameter deviations are known to be favorable
for enhanced hurricane activity.
27
Figure 11: Linear correlations between May 925 mb zonal wind in the central tropical
Pacific (Predictor 2) and August-October sea surface temperature (panel a), AugustOctober sea level pressure (panel b), August-October 925 mb zonal wind (panel c) and
August-October 200 mb zonal wind (panel d). All four of these parameter deviations are
known to be favorable for enhanced hurricane activity. Zonal wind values have been
multiplied by -1 to allow for easy comparison with Figure 1.
28
Figure 12: Linear correlations between April-May 200 mb meridional wind in the
Caribbean (Predictor 3) and August-October sea surface temperature (panel a), AugustOctober sea level pressure (panel b), August-October 925 mb zonal wind (panel c) and
August-October 200 mb zonal wind (panel d). All four of these parameter deviations are
known to be favorable for enhanced hurricane activity.
5
Analog-Based Predictors for 2010 Hurricane Activity
Certain years in the historical record have global oceanic and atmospheric trends
which are substantially similar to 2010. These years also provide useful clues as to likely
29
trends in activity that the forthcoming 2010 hurricane season may bring. For this early
June extended range forecast, we determine which of the prior years in our database have
distinct trends in key environmental conditions which are similar to current April-May
2010 conditions. Table 10 lists our analog selections.
We select prior hurricane seasons since 1949 which have similar atmosphericoceanic conditions to those currently being experienced. We searched for years that were
generally characterized by weak El Niño or neutral conditions, well above-average
tropical Atlantic SSTs and above-average far North Atlantic SSTs during April-May.
There were four hurricane seasons since 1949 with characteristics most similar to
what we observed in April-May 2010. The best analog years that we could find for the
2010 hurricane season were 1958, 1966, 1969, and 2005. We anticipate that 2010
seasonal hurricane activity will have activity in line with what was experienced in the
average of these four years. We believe that 2010 will thus be a very active hurricane
season.
Table 10: Best analog years for 2010 with the associated hurricane activity listed for
each year.
Year
1958
1966
1969
2005
Mean
2010 Forecast
6
NS
10
11
18
28
16.8
NSD
55.50
64.00
91.50
131.50
85.6
H
7
7
12
15
10.3
HD
30.25
41.75
40.00
49.75
40.4
MH
5
3
5
7
5.0
MHD
9.50
8.75
6.75
17.75
10.7
ACE
121
145
166
250
171
NTC
144
140
182
279
187
18
90
10
40
5
13
185
195
ENSO
Moderate-to-strong El Niño conditions were in place during the winter of 20092010. These conditions have weakened significantly over the past several months. SSTs
are generally about average across the eastern and central tropical Pacific, indicating that
El Niño has transitioned to neutral conditions. Table 11 displays March and May SST
anomalies for several Nino regions. Note that all regions except for Nino 1+2 have
experienced considerable cooling since March.
30
Table 11: March and May SST anomalies for Nino 1+2, Nino 3, Nino 3.4, and Nino 4,
respectively. May-March SST anomaly differences are also provided.
Region
Nino 1+2
Nino 3
Nino 3.4
Nino 4
March SST
Anomaly (°C)
-0.2
0.7
1.1
1.1
May SST
Anomaly (°C)
0.0
0.0
-0.1
0.4
May – March
SST Anomaly (°C)
+0.2
-0.7
-1.2
-0.7
We have increased confidence that we will have weak La Niña conditions
(August-October-averaged Nino 3.4 between -1.0° and -0.5°C) during the peak of this
year’s hurricane season. We are approaching the end of the ENSO springtime
predictability barrier, and therefore, statistical and dynamical models of ENSO show
improved skill from a May initialization period compared with models initialized earlier
in the spring. All statistical and dynamical models are calling for either neutral
conditions (Nino 3.4 SSTs between -0.5°C and 0.5°C) or a transition to La Niña by the
August-October period (Nino 3.4 SSTs < -0.5°C) (Figure 13). In general, the dynamical
models are calling for a greater likelihood of La Niña conditions by August-October,
while the statistical models tend to maintain more neutral conditions. We believe that we
will experience a weak La Niña event by September.
We find that, in general, the European Centre for Medium-Range Weather
Forecasts (ECMWF) shows the best prediction skill of the various ENSO models that are
currently available. The ECMWF model is calling for an August-October-averaged Nino
3.4 SST anomaly of -0.9°C, giving us increased confidence that El Niño conditions will
not have a detrimental role on this year’s hurricane season as they did during last year’s
hurricane season. Note that none of the ECMWF ensemble members have El Niño
conditions beyond July, with most members calling for either a weak or moderate La
Niña by September.
31
Figure 13: ENSO forecasts from various statistical and dynamical models. Figure
courtesy of the International Research Institute (IRI). All forecast models are calling for
either neutral or La Niña conditions during August-October.
32
Figure 14: ECMWF ensemble model forecast for the Nino 3.4 region. All ensemble
members predict below-average SSTs in the Nino 3.4 region by September, with most
ensemble members calling for either a weak or moderate La Niña event.
Another reason why we believe that weak La Niña conditions will be present during this
year’s hurricane season is due to the current sub-surface ocean temperature anomaly
pattern. Note that below-normal equatorial heat anomalies have reached the eastern
Pacific. This reduction in upper-ocean heat content is typically seen with the transition
from a weak El Niño to neutral conditions (Figure 15).
33
Figure 15: Current equatorial sub-surface upper-ocean (top 300 meters) heat
anomaly (°C) pattern in the tropical Pacific.
Based on this information, we believe that the transition to neutral conditions will
continue for the next couple of months with weak La Niña conditions likely developing
thereafter. El Niños typically increase levels of vertical wind shear in the tropical
34
Atlantic, causing detrimental conditions for Atlantic tropical cyclone formation and
intensification. Since we do not expect El Niño this year, we do not expect to see the
high levels of vertical shear across the Main Development Region that we experienced
last year.
7
Current Atlantic Basin Conditions
Conditions in the Atlantic remain quite favorable for an active season. SST
anomalies across the Main Development Region for May are near their highest levels on
record. Figure 16 displays the currently-observed SST anomaly pattern across the
Atlantic. Note the very strong positive anomalies throughout the tropical Atlantic and the
cool anomalies in the Gulf of Mexico and off the East Coast of the United States. This
SST anomaly pattern is characteristic of the negative North Atlantic Oscillation (NAO)
index values that have tended to be present over the past several months. A negative
NAO is characterized by anomalous high pressure in the northern Atlantic and anomalous
low pressure near the Azores High (Figure 17). This pressure gradient pattern causes a
reduction in the trade winds across the tropical Atlantic (Figure 18). Reduced trade
winds drive less upwelling and evaporation from the sea surface, typically resulting in a
warming of SSTs. Warmer Atlantic SSTs in the MDR are associated with an active
THC/positive AMO, weaker tropospheric vertical wind shear, weaker trade winds,
increased instability and lower-than-normal sea level pressures. All of these conditions
are generally associated with much more active Atlantic basin hurricane seasons.
Another parameter that we are currently monitoring is the amount of dust being
advected off of the west coast of North Africa from the Sahara. The correlation between
May Saharan Air Layer (SAL) dust loadings across the Main Development Region and
NTC is -0.40 based on the period from 1982-2008. The dust loadings from the SAL this
May have been approximately 0.5 standard deviations above average across the Main
Development Region; however, the current outlook for dust across the Main
Development Region for the 2010 hurricane season is for average to below-average dust
content, another slightly enhancing factor for an active hurricane season (Amato Evan,
2010, personal communication).
35
Figure 16: May 2010 SST anomaly (°C) pattern across the Atlantic Ocean.
Figure 17: Sea level pressure anomaly (mb) pattern observed during April-May 2010.
36
Figure 18: Anomalous low-level wind (ms-1) observed during April-May 2010.
Westerly anomalies were observed over the tropical Atlantic, indicating a reduction in the
easterly trade winds. Weaker trades typically warm the tropical Atlantic.
8
Adjusted 2010 Forecast
Table 12 shows our final adjusted early June forecast for the 2010 season which is
a combination of our statistical scheme, our analog forecast and qualitative adjustments
for other factors not explicitly contained in any of these schemes. Our statistical forecast
calls for a moderately active season, while our analog forecast indicates activity at well
above-average levels. We favor our analog forecast technique with this particular
prediction given the rapid cooling in the eastern and central tropical Pacific. We foresee
a very active Atlantic basin hurricane season due to currently-observed very favorable
conditions in the tropical Atlantic and the recent demise of El Niño.
37
Table 12: Summary of our early June statistical forecast, our analog forecast and our
adjusted final forecast for the 2010 hurricane season.
Forecast Parameter and 1950-2000 Climatology
(in parentheses)
Named Storms (9.6)
Named Storm Days (49.1)
Hurricanes (5.9)
Hurricane Days (24.5)
Major Hurricanes (2.3)
Major Hurricane Days (5.0)
Accumulated Cyclone Energy Index (96.1)
Net Tropical Cyclone Activity (100%)
9
Statistical
Scheme
10.5
52.5
6.2
25.3
2.7
6.5
103
112
Analog
Scheme
16.8
85.6
10.3
40.4
5.0
10.7
171
187
Adjusted Final
Forecast
18
90
10
40
5
13
185
195
Landfall Probabilities for 2010
A significant focus of our recent research involves efforts to develop forecasts of
the probability of hurricane landfall along the U.S. coastline and in the Caribbean.
Whereas individual hurricane landfall events cannot be accurately forecast months in
advance, the total seasonal probability of landfall can be forecast with statistical skill.
With the observation that, statistically, landfall is a function of varying climate
conditions, a probability specification has been developed through statistical analyses of
all U.S. hurricane and named storm landfall events during the 20th century (1900-1999).
Specific landfall probabilities can be given for all tropical cyclone intensity classes for a
set of distinct U.S. coastal regions.
Net landfall probability is shown linked to the overall Atlantic basin Net Tropical
Cyclone activity (NTC; see Table 13). NTC is a combined measure of the year-to-year
mean of six indices of hurricane activity, each expressed as a percentage difference from
the long-term average. Long-term statistics show that, on average, the more active the
overall Atlantic basin hurricane season is, the greater the probability of U.S. hurricane
landfall.
Table 13: NTC activity in any year consists of the seasonal total of the following six
parameters expressed in terms of their long-term averages. A season with 10 NS, 50
NSD, 6 H, 25 HD, 3 MH, and 5 MHD would then be the sum of the following ratios:
10/9.6 = 104, 50/49.1 = 102, 6/5.9 = 102, 25/24.5 = 102, 3/2.3 = 130, 5/5.0 = 100,
divided by six, yielding an NTC of 107.
1)
2)
3)
4)
5)
6)
1950-2000 Average
Named Storms (NS)
Named Storm Days (NSD)
Hurricanes (H)
Hurricane Days (HD)
Major Hurricanes (MH)
Major Hurricane Days (MHD)
38
9.6
49.1
5.9
24.5
2.3
5.0
Table 14 lists strike probabilities for the 2010 hurricane season for different TC
categories for the entire U.S. coastline, the Gulf Coast and the East Coast including the
Florida peninsula. In our early June forecast of 2009, we initiated landfall probabilities
for various islands and landmasses in the Caribbean and in Central America. Note that
Atlantic basin NTC activity in 2010 is expected to be above its long-term average of 100,
and therefore, landfall probabilities are above their long-term average.
Please visit the Landfalling Probability Webpage at http://www.etransit.org/hurricane for landfall probabilities for 11 U.S. coastal regions and 205 coastal
and near-coastal counties from Brownsville, Texas to Eastport, Maine. The probability
of each U.S. coastal state being impacted by hurricanes and major hurricanes is also
included. In addition, we now include probabilities of named storms, hurricanes and
major hurricanes tracking within 50 and 100 miles of various islands and landmasses in
the Caribbean and Central America.
Table 14: Estimated probability (expressed in percent) of one or more landfalling tropical
storms (TS), category 1-2 hurricanes (HUR), category 3-4-5 hurricanes, total hurricanes
and named storms along the entire U.S. coastline, along the Gulf Coast (Regions 1-4),
and along the Florida Peninsula and the East Coast (Regions 5-11) for 2010.
Probabilities of a tropical storm, hurricane and major hurricane tracking into the
Caribbean are also provided. The long-term mean annual probability of one or more
landfalling systems during the last 100 years is given in parentheses.
Region
Entire U.S. (Regions 1-11)
Gulf Coast (Regions 1-4)
Florida plus East Coast (Regions 5-11)
Caribbean (10-20°N, 60-88°W)
TS
95% (79%)
82% (59%)
74% (50%)
97% (82%)
Category 1-2
HUR
89% (68%)
66% (42%)
68% (44%)
81% (57%)
Category 3-4-5
HUR
76% (52%)
50% (30%)
51% (31%)
65% (42%)
All
HUR
97% (84%)
83% (60%)
84% (61%)
93% (75%)
Named
Storms
99% (97%)
97% (83%)
96% (81%)
99% (96%)
10
Have Atmospheric CO2 Increases Been Responsible for the
Recent Large Upswing (since 1995) in Atlantic Basin Major
Hurricanes?
A. BACKGROUND
The U.S. landfall of major hurricanes Dennis, Katrina, Rita and Wilma in 2005 and the
four Southeast landfalling hurricanes of 2004 – Charley, Frances, Ivan and Jeanne, raised
questions about the possible role that global warming played in those two unusually
destructive seasons. In addition, three category 2 hurricanes (Dolly, Gustav and Ike)
pummeled the Gulf Coast in 2008 causing considerable devastation. Some researchers
have tried to link the rising CO2 levels with SST increases during the late 20th century
and say that this has brought on higher levels of hurricane intensity.
39
These speculations that hurricane intensity has increased have been given much media
attention; however, we believe that they are not valid, given current observational data.
There has, however, been a large increase in Atlantic basin major hurricane activity since
1995 in comparison with the prior 15-year period of 1980-1994 (Figure 19) and the prior
quarter-century period of 1970-1994. It has been tempting for many who do not have a
strong background in hurricane knowledge to jump on this recent 15-year increase in
major hurricane activity as strong evidence of a human influence on hurricanes. It should
be noted, however, that the last 15-year active major hurricane period of 1995-2009 has,
however, not been more active than the earlier 15-year period of 1950-1964 when the
Atlantic Ocean circulation conditions were similar to what has been observed in the last
15 years. These conditions occurred even though atmospheric CO2 amounts were lower
in the earlier period.
Figure 19: The tracks of major (Category 3-4-5) hurricanes during the 15-year period of
1995-2009 when the Atlantic thermohaline circulation (THC) was strong versus the prior
15-year period of 1980-1994 when the THC was weak. Note that there were more than
2.5 times as many major hurricanes when the THC was strong as when it was weak.
Table 15 shows how large Atlantic basin hurricane variations are between strong and
weak THC periods. Note especially how large the ratio is for major hurricane days (3.8)
during strong vs. weak THC periods. Normalized U.S. hurricane damage studies by
Pielke and Landsea (1998) show that landfalling major hurricanes account on average for
about 80-85 percent of all hurricane-related destruction even though these major
hurricanes make up only 20-25 percent of named storms.
Although global surface temperatures increased during the late 20th century, there is no
reliable data to indicate increased hurricane frequency or intensity in any of the globe’s
other tropical cyclone basins since 1979. Global Accumulated Cyclone Energy (ACE)
40
shows significant year-to-year and decadal variability over the past thirty years but no
increasing trend (Figure 20). Similarly, Klotzbach (2006) found no significant change in
global TC activity during the period from 1986-2005.
Table 15: Comparison of Atlantic basin hurricane activity in two 15-year periods when
the Atlantic Ocean THC (or AMO) was strong versus an intermediate period (1970-1994)
when it was weak.
THC
SST (10-15oN;
70-40oW)
Avg.
CO2
ppm
NS
NSD
H
HD
MH
MHD
ACE
NTC
1950-1964 (15
years)
Strong
27.93
320
9.9
53.6
6.5
30.5
3.8
9.8
122
134
1970-1994 (25
years)
Weak
27.60
345
9.3
41.9
5.0
16.0
1.5
2.5
68
75
1995-2009 (15
years)
Strong
27.97
372
14.5
73.9
7.7
31.9
3.7
9.2
140
151
∆ 0.35oC
~0
1.3
1.5
1.4
1.9
2.5
3.8
1.9
1.9
Year Ratio
Strong/Weak THC
Figure 20: Northern Hemisphere, Southern Hemisphere, and global ACE over the period
from 1979-2009. Figure has been adapted from Ryan Maue, Center for OceanAtmospheric Prediction Studies, Florida State University.
41
Causes of Upswing in Atlantic Major Hurricane Activity since 1995. The Atlantic
Ocean has a strong multi-decadal signal in its hurricane activity which is likely due to
multi-decadal variations in the strength of the Atlantic Ocean thermohaline circulation
(THC) (Figure 21). The oceanic and atmospheric response to the THC is often referred
to as the Atlantic Multi-decadal Oscillation (AMO). We use the THC and AMO
interchangeably throughout the remainder of this discussion. The strength of the THC
can never be directly measured, but it can be diagnosed, as we have done, from the
magnitude of the sea surface temperature anomaly (SSTA) in the North Atlantic (Figure
22) combined with the sea level pressure anomaly (SLPA) in the Atlantic between the
latitude of the equator and 50°N (Klotzbach and Gray 2008).
The THC (or AMO) is strong when there is an above-average poleward advection of
warm low-latitude waters to the high latitudes of the North Atlantic. This water can then
sink to deep levels when it reaches the far North Atlantic in a process known as deep
water formation. The water then moves southward at deep levels in the ocean. The
amount of North Atlantic water that sinks is proportional to the water’s density which is
determined by its salinity content as well as its temperature. Salty water is denser than
fresh water at water temperatures near freezing. There is a strong association between
North Atlantic SSTA and North Atlantic salinity (Figure 23). High salinity implies
higher rates of North Atlantic deep water formation (or subsidence) and thus a stronger
flow of upper level warm water from lower latitudes as replacement. See the papers by
Gray et al. (1999), Goldenberg et al. (2001), and Grossman and Klotzbach (2009) for
more discussion.
Figure 21: Illustration of strong (top) and weak (bottom) phases of the THC or AMO.
42
Figure 22: Long-period portrayal (1878-2006) of North Atlantic sea surface temperature
anomalies (SSTA). The red (warm) periods are when the THC (or AMO) is stronger than
average and the blue periods are when the THC (or AMO) is weaker than average.
Figure 23: Illustration of the strong association of yearly average North Atlantic SSTA
and North Atlantic salinity content between 1958 and 2004.
B. WHY CO2 INCREASES ARE NOT RESPONSIBLE FOR ATLANTIC SST AND
HURRICANE ACTIVITY INCREASES
43
Theoretical considerations do not support a close relationship between SSTs and
hurricane intensity. In a global warming world, the atmosphere’s upper air temperatures
will warm or cool in unison with longer-period SST changes. Vertical lapse rates will
thus not be significantly altered in a somewhat warmer or somewhat cooler tropical
oceanic environment. We have no plausible physical reasons for believing that Atlantic
hurricane frequency or intensity will significantly change if global or Atlantic Ocean
temperatures were to rise by 1-2°C. Without corresponding changes in many other basic
features, such as vertical wind shear or mid-level moisture, little or no additional TC
activity should occur with SST increases.
Confusing Time Scales of SST Influences. A hurricane passing over a warmer body of
water, such as the Gulf Stream, will often undergo some intensification. This is due to
the sudden lapse rate increase which the hurricane’s inner core experiences when it
passes over warmer water. The warmer SSTs cause the hurricane’s lower boundary layer
temperature and moisture content to rise. While these low level changes are occurring,
upper tropospheric conditions are often not altered significantly. These rapidly occurring
lower- and upper-level temperature differences cause the inner-core hurricane lapse rates
to increase and produce more intense inner-core deep cumulus convection. This typically
causes a rapid increase in hurricane intensity. Such observations have led many
observers to directly associate SST increases with greater hurricane intensity potential.
This is valid reasoning for day-to-day hurricane intensity change associated with
hurricanes moving over warmer or colder patches of SST. But such direct reasoning does
not hold for conditions occurring in an overall climatologically warmer (or cooler)
tropical oceanic environment where broad-scale global and tropical rainfall conditions are
not expected to significantly vary. During long-period climate change, temperature and
moisture conditions rise at both lower and upper levels. Lapse rates are little effected
(Figure 24).
44
Figure 24: Illustration of how SST increases of 1°C will bring about higher planetary
boundary layer (PBL) temperature and moisture increases that will also occur in small
amounts throughout the troposphere. The combination of these changes is such that
potential buoyancy for cumulonimbus (Cb) development is not much altered by increases
in SST alone.
Any warming-induced increase in boundary layer temperature and moisture will be (to
prevent significant global rainfall alteration) largely offset by a similar but weaker change
through the deep troposphere up to about 10 km height. Upper-tropospheric changes are
weaker than boundary layer changes, but they occur through a much deeper layer. These
weaker and deeper compensating increases in upper-level temperature and moisture are
necessary to balance out the larger increases in temperature and moisture which occur in
the boundary layer. Global and tropical rainfall would be altered significantly if broadscale lapse rates were ever altered to an appreciable degree.
Thus, we cannot automatically assume that with warmer global SSTs that we will
necessarily have more intense hurricanes due to lapse-rate alterations. We should not
expect that the frequency and/or intensity of Category 4-5 hurricanes will necessarily
change as a result of changes in global or individual storm basin SSTs. Historical
evidence does not support hurricanes being less intense during the late 19th century and
the early part of the 20th century when SSTs were slightly lower.
CO2 Influence on Hurricane Activity. We have been performing research with the
International Satellite Cloud Climatology Project (ISCCP) and the NOAA National
Centers for Environmental Prediction (NCEP) Reanalysis data sets. We have used this
data to make an annual average of the global tropical (30°N-30°S; 0-360°) energy budget
(Figure 25) for the years from 1984-2004. Note that the various surface and top of the
atmosphere energy fluxes are very large. For the tropical surface, for instance, there are
637 Wm-2 units of downward incoming solar and infrared (IR) energy. This downward
energy flux is largely balanced by an upward surface energy flux of 615 Wm-2 which is
45
due to upward fluxes from IR radiation, evaporated liquid water, and sensible heat.
Similar large energy fluxes are present at the top of the atmosphere and within the
troposphere.
Figure 25: Vertical cross-section of the annual tropical energy budget as determined
from a combination of ISCCP and NCEP Reanalysis data over the period of 1984-2004.
Abbreviations are IR for longwave infrared radiation, Alb for albedo and OLR for
outgoing longwave radiation. The tropics receive an excess of about 44 Wm-2 radiation
energy which is convected and exported as sensible heat to latitudes poleward of 30°.
Estimates are about half (22 Wm-2) of this excess is transported by the atmosphere and
the other half is transported by the oceans. Note, on the right, how small an OLR
blockage has occurred up to now due to CO2 increases (~ 1.3 Wm-2) and the blockage of
3.7 Wm-2 that will occur from a doubling of CO2 by the end of this century.
It has been estimated that a doubling of CO2 (from the pre-industrial period) without any
feedback influences would result in a blockage of OLR to space of about 3.7 Wm-2. The
currently-measured value of CO2 in the atmosphere is 380 parts per million by volume
(ppmv). If we take the background pre-industrial value of CO2 to be 280 ppmv, then by
theory we should currently be having (from CO2 increases alone) about (100/280)*3.7 =
1.3 Wm-2 less OLR energy flux to space than was occurring in the mid-19th century.
This reduced OLR of 1.3 Wm-2 is very small in comparison with most of the other
tropical energy budget exchanges. Slight changes in any of these other larger tropical
energy budget components could easily negate or reverse this small CO2-induced OLR
blockage. For instance, an upper tropospheric warming of about 1°C with no change in
moisture would enhance OLR sufficiently that it would balance the reduced OLR
influence from a doubling of CO2. Similarly, if there were a reduction of upper level
water vapor such that the long wave radiation emission level to space were lowered about
46
7 mb (~ 140 m) there would be an enhancement of OLR (with no change of temperature)
sufficient to balance the suppression of OLR from a doubling of CO2. The 1.3 Wm-2
reduction in OLR we have experienced since the mid-19th century (about one-third of the
way to a doubling of CO2) is very small compared with the overall 399 Wm-2 of solar
energy impinging on the top of the tropical atmosphere and the mostly compensating 356
Wm-2 of OLR and albedo energy going back to space. This 1.3 Wm-2 energy gain is
much too small to ever allow a determination of its possible influence on TC activity.
Any such potential CO2 influence on TC activity is deeply buried as turbulence within the
tropical atmospheres’ many other energy components. It is possible that future higher
atmospheric CO2 levels may cause a small influence on global TC activity. But any such
potential influence would likely never be able to be detected, given that our current
measurement capabilities only allow us to assess TC intensity to within about 5 mph.
C. CONTRAST OF THEORIES OF HURRICANE ACTIVITY CHANGES
Theory of Human-Induced Increases due to Rising CO2 Levels. Those who think
CO2 increases have and will cause significant increases in hurricane activity believe that
the physics of the CO2-hurricane association is directly related to radiation changes as
indicated in Figure 26. They view CO2 as blocking OLR to space. This acts to warm
SSTs and add moisture to the boundary layer just above the ocean surface. These
changes cause an increase in lapse rates (the lower levels warm while upper levels do not
change much) which lead to more deep cumulonimbus convection (Cb). More Cb
convection leads to a higher percentage of tropical disturbances forming into tropical
cyclones and a greater spin-up of the inner-core of those systems which do form.
Figure 26: Physical linkage of those who believe that increases in CO2 are making
hurricanes more frequent and/or more intense.
This physical argument is too simplistic. It has no empirical verification in any other
global TC basin except for the Atlantic. Table 16 shows the correlation of ACE with late
47
summer-early fall SSTs in the Northeast Pacific, the Northwest Pacific and the Southern
Hemisphere. Note the low (or even negative) correlations between ACE and SST in each
of these three TC basins. It is obvious that other physical processes besides SST are
primarily responsible for differences in hurricane activity in these basins.
Table 16: Correlation of ACE with late summer-early fall SSTs in three TC basins from
1980-2009.
Yearly Mean ACE
ACE vs. SST Correlation (r)
Northeast Pacific
134
0.01
Northwest Pacific
310
-0.30
Southern Hemisphere
205
0.23
Globe
(SST 20oN-20oS)
769
-0.08
Theory of the THC (or AMO). We do not view seasonal hurricane variability in the
Atlantic as being directly related to changes in CO2-induced radiation forcing or to SST
changes by themselves. For the Atlantic, we view long-period tropical cyclone
variability primarily as a result of changes in the strength of the THC (or AMO). We
hypothesize that these changes act as shown in Figure 27 and as discussed in the previous
section. THC changes result in altering tropospheric vertical wind shear, trade wind
strength, and SSTs in the Main Development Region (MDR) of 10-20°N; 20-70°W in the
tropical Atlantic. A large component of the SST increase in this area is not a direct result
of radiation differences but rather the combination of the effects of reduced southward
advection of colder water in the east Atlantic and reduced trade wind strength. Weaker
trade winds reduce upwelling and evaporation and typically act to increase SST.
The influence of the warmer Atlantic SST, as previously discussed, is not primarily to
enhance lapse rates and Cb convection but rather as a net overall positive influence on
lowering the MDR’s surface pressure and elevating mean upward tropospheric vertical
motion. This causes an increase in tropospheric moisture content.
48
Figure 27: Idealized portrayal of how changes in the Atlantic THC bring about various
parameter changes in the Atlantic’s MDR. Vertical shear, trade wind strength and SST
are the key parameters which respond to THC changes. Favorable SLP and mid-level
moisture changes occur in association with the shear, trade wind and SST changes.
D. DISCUSSION
In a global warming or global cooling world, the atmosphere’s upper air temperatures
will warm or cool in unison with the SSTs. Vertical lapse rates will not be significantly
altered. We have no plausible physical reasons for believing that Atlantic hurricane
frequency or intensity will change significantly if global ocean temperatures were to
continue to rise. For instance, in the quarter-century period from 1945-1969 when the
globe was undergoing a weak cooling trend, the Atlantic basin experienced 80 major (Cat
3-4-5) hurricanes and 201 major hurricane days. By contrast, in a similar 25-year period
from 1970-1994 when the globe was undergoing a general warming trend, there were
only 38 Atlantic major hurricanes (48% as many) and 63 major hurricane days (31% as
many) (Figure 28). Atlantic SSTs and hurricane activity do not follow global mean
temperature trends.
49
Figure 28: Tracks of major (Category 3-4-5) hurricanes during the 25-year period of
1945-1969 when the globe was undergoing a weak cooling versus the 25-year period of
1970-1994 when the globe was undergoing a modest warming. CO2 amounts in the later
period were approximately 18 percent higher than in the earlier period. Major Atlantic
hurricane activity was only about one-third as frequent during the latter period despite
warmer global temperatures.
The most reliable long-period hurricane records we have are the measurements of US
landfalling TCs since 1900 (Table 17). Although global mean ocean and Atlantic SSTs
have increased by about 0.4°C between these two 55-year periods (1900-1954 compared
with 1955-2009), the frequency of US landfall numbers actually shows a slight
downward trend for the later period. This downward trend is particularly noticeable for
the US East Coast and Florida Peninsula where the difference in landfall of major
(Category 3-4-5) hurricanes between the 44-year period of 1922-1965 (24 landfall
events) and the 44-year period of 1966-2009 (7 landfall events) was especially large
(Figure 29). For the entire United States coastline, 38 major hurricanes made landfall
during the earlier 44-year period (1922-1965) compared with only 26 major hurricanes
for the latter 44-year period (1966-2009). This occurred despite the fact that CO2
averaged approximately 365 ppm during the latter period compared with 310 ppm during
the earlier period.
50
Table 17: U.S. landfalling tropical cyclones by intensity during two 55-year periods.
YEARS
Named Storms
Hurricanes
Major
Hurricanes
(Cat 3-4-5)
1900-1954 (55 years)
208
113
44
Global
Temperature
Increase
+0.4°C
1955-2009 (55 years)
184
90
36
We should not read too much into the three very active hurricane seasons of 2004, 2005,
and 2008. The activity of these years was unusual but well within natural bounds of
hurricane variation.
What made the 2004-2005 and 2008 seasons so destructive was not the high frequency of
major hurricanes but the high percentage of hurricanes that were steered over the US
coastline. The US hurricane landfall events of these years were primarily a result of the
favorable upper-air steering currents present during these years.
Figure 29: Contrast of tracks of East Coast and Florida Peninsula major landfalling
hurricanes during the 44-year period of 1922-1965 versus the most recent 44-year period
of 1966-2009.
51
Although 2005 had a record number of TCs (28 named storms), this should not be taken
as an indication of something beyond natural processes. There have been several other
years with comparable hurricane activity to 2005. For instance, 1933 had 21 named
storms in a year when there was no satellite or aircraft data. Records of 1933 show all 21
named storms had tracks west of 60°W where surface observations were more plentiful.
If we eliminate all of the named storms of 2005 whose tracks were entirely east of 60°W
and therefore may have been missed given the technology available in 1933, we reduce
the 2005 named storm total by seven (to 21) – the same number as was observed to occur
in 1933.
Utilizing the National Hurricane Center’s best track database of hurricane records back to
1875, six previous seasons had more hurricane days than the 2005 season. These years
were 1878, 1893, 1926, 1933, 1950 and 1995. Also, five prior seasons (1893, 1926,
1950, 1961 and 2004) had more major hurricane days. Although the 2005 hurricane
season was certainly one of the most active on record, it was not as much of an outlier as
many have indicated.
We believe that the Atlantic basin remains in an active hurricane cycle associated with a
strong THC. This active cycle is expected to continue for another decade or two at which
time we should enter a quieter Atlantic major hurricane period like we experienced
during the quarter-century periods of 1970-1994 and 1901-1925. Atlantic hurricanes go
through multi-decadal cycles. Cycles in Atlantic major hurricanes have been
observationally traced back to the mid-19th century. Changes in the THC (or AMO) have
been inferred from Greenland paleo ice-core temperature measurements going back
thousand of years. These changes are natural and have nothing to do with human
activity.
11
Forthcoming Updated Forecasts of 2010 Hurricane Activity
We will be issuing a seasonal update of our 2010 Atlantic basin hurricane forecast
on Wednesday 4 August. We will also be issuing two-week forecasts for Atlantic TC
activity during the climatological peak of the season from August-October. A
verification and discussion of all 2010 forecasts will be issued in late November 2010.
Our first seasonal hurricane forecast for the 2011 hurricane season will be issued in early
December 2010. All of these forecasts will be available on the web at:
http://hurricane.atmos.colostate.edu/Forecasts.
12
Acknowledgments
Besides the individuals named on page 5, there have been a number of other
meteorologists that have furnished us with data and given valuable assessments of the
current state of global atmospheric and oceanic conditions. These include Brian
McNoldy, Art Douglas, Ray Zehr, Mark DeMaria, Todd Kimberlain, Paul Roundy and
Amato Evan and Ethan Gibney. In addition, Barbara Brumit and Amie Hedstrom have
provided excellent manuscript, graphical and data analysis and assistance over a number
of years. We have profited over the years from many in-depth discussions with most of
52
the current and past NHC hurricane forecasters. The second author would further like to
acknowledge the encouragement he has received for this type of forecasting research
application from Neil Frank, Robert Sheets, Robert Burpee, Jerry Jarrell, and Max
Mayfield, former directors of the National Hurricane Center (NHC) as well as Bill Read,
current director of the NHC. Uma Shama, Larry Harman and Daniel Fitch of
Bridgewater State College, MA have provided assistance and technical support in the
development of our Landfalling Hurricane Probability Webpage.
The financial backing for the issuing and verification of these forecasts has been
supported in part by the National Science Foundation. We also thank the GeoGraphics
Laboratory at Bridgewater State College for their assistance in developing the
Landfalling Hurricane Probability Webpage.
53
13
Citations and Additional Reading
Blake, E. S., 2002: Prediction of August Atlantic basin hurricane activity. Dept. of Atmos. Sci. Paper No.
719, Colo. State Univ., Ft. Collins, CO, 80 pp.
Blake, E. S. and W. M. Gray, 2004: Prediction of August Atlantic basin hurricane activity. Wea.
Forecasting, 19, 1044-1060.
Chiang, J. C. H. and D. J. Vimont, 2004: Analogous Pacific and Atlantic meridional modes of tropical
atmosphere-ocean variability. J. Climate, 17, 4143-4158.
Compo, G. P., J. S. Whitaker, and P. D. Sardeshmukh, 2006: Feasibility of a 100 year reanalysis using only
surface pressure data. Bull. Amer. Meteor. Soc., 87, 175-190.
DeMaria, M., J. A. Knaff and B. H. Connell, 2001: A tropical cyclone genesis parameter for the tropical
Atlantic. Wea. Forecasting, 16, 219-233.
Elsner, J. B., G. S. Lehmiller, and T. B. Kimberlain, 1996: Objective classification of Atlantic hurricanes.
J. Climate, 9, 2880-2889.
Evan, A. T., J. Dunion, J. A. Foley, A. K. Heidinger, and C. S. Velden, 2006: New evidence for a
relationship between Atlantic tropical cyclone activity and African dust outbreaks, Geophys. Res.
Lett, 33, doi:10.1029/2006GL026408.
Goldenberg, S. B., C. W. Landsea, A. M. Mestas-Nunez, and W. M. Gray, 2001: The recent increase in
Atlantic hurricane activity: Causes and Implications. Science, 293, 474-479.
Goldenberg, S. B. and L. J. Shapiro, 1996: Physical mechanisms for the association of El Niño and West
African rainfall with Atlantic major hurricane activity. J. Climate, 1169-1187.
Gray, W. M., 1984a: Atlantic seasonal hurricane frequency: Part I: El Niño and 30 mb quasi-biennial
oscillation influences. Mon. Wea. Rev., 112, 1649-1668.
Gray, W. M., 1984b: Atlantic seasonal hurricane frequency: Part II: Forecasting its variability. Mon. Wea.
Rev., 112, 1669-1683.
Gray, W. M., 1990: Strong association between West African rainfall and US landfall of intense
hurricanes. Science, 249, 1251-1256.
Gray, W. M., and P. J. Klotzbach, 2003 and 2004: Forecasts of Atlantic seasonal and monthly hurricane
activity and US landfall strike probability. Available online at http://hurricane.atmos.colostate.edu
Gray, W. M., C. W. Landsea, P. W. Mielke, Jr., and K. J. Berry, 1992: Predicting Atlantic seasonal
hurricane activity 6-11 months in advance. Wea. Forecasting, 7, 440-455.
Gray, W. M., C. W. Landsea, P. W. Mielke, Jr., and K. J. Berry, 1993: Predicting Atlantic basin seasonal
tropical cyclone activity by 1 August. Wea. Forecasting, 8, 73-86.
Gray, W. M., C. W. Landsea, P. W. Mielke, Jr., and K. J. Berry, 1994a: Predicting Atlantic basin seasonal
tropical cyclone activity by 1 June. Wea. Forecasting, 9, 103-115.
Gray, W. M., J. D. Sheaffer and C. W. Landsea, 1996: Climate trends associated with multi-decadal
variability of intense Atlantic hurricane activity. Chapter 2 in “Hurricanes, Climatic Change and
54
Socioeconomic Impacts: A Current Perspective", H. F. Diaz and R. S. Pulwarty, Eds., Westview
Press, 49 pp.
Gray, W. M., 1998: Atlantic Ocean influences on multi-decadal variations in El Niño frequency and
intensity. Ninth Conference on Interaction of the Sea and Atmosphere, 78th AMS Annual
Meeting, 11-16 January, Phoenix, AZ, 5 pp.
Grossmann, I., and P. J. Klotzbach, 2009: A review of North Atlantic modes of natural variability and their
driving mechanisms. J. Geophys Res., 114, D24107, doi: 10.1029/2009JD012728.
Henderson-Sellers, A., H. Zhang, G. Berz, K. Emanuel, W. Gray, C. Landsea, G. Holland, J. Lighthill, S-L.
Shieh, P. Webster, and K. McGuffie, 1998: Tropical cyclones and global climate change: A postIPCC assessment. Bull. Amer. Meteor. Soc., 79, 19-38.
Klotzbach, P. J., 2002: Forecasting September Atlantic basin tropical cyclone activity at zero and onemonth lead times. Dept. of Atmos. Sci. Paper No. 723, Colo. State Univ., Ft. Collins, CO, 91 pp.
Klotzbach, P. J., 2006: Trends in global tropical cyclone activity over the past twenty years (1986-2005).
Geophys. Res. Lett., 33, doi:10.1029/2006GL025881.
Klotzbach, P. J., 2007: Revised prediction of seasonal Atlantic basin tropical cyclone activity from 1
August. Wea. and Forecasting, 22, 937-949.
Klotzbach, P. J. and W. M. Gray, 2003: Forecasting September Atlantic basin tropical cyclone activity.
Wea. and Forecasting, 18, 1109-1128.
Klotzbach, P. J. and W. M. Gray, 2004: Updated 6-11 month prediction of Atlantic basin seasonal
hurricane activity. Wea. and Forecasting, 19, 917-934.
Klotzbach, P. J. and W. M. Gray, 2006: Causes of the unusually destructive 2004 Atlantic basin hurricane
season. Bull. Amer. Meteor. Soc., 87, 1325-1333.
Klotzbach, P. J., and W. M. Gray, 2008: Multidecadal variability in North Atlantic tropical cyclone
activity. J. Climate, 21, 3929-3935.
Knaff, J. A., 1997: Implications of summertime sea level pressure anomalies. J. Climate, 10, 789-804.
Knaff, J. A., 1998: Predicting summertime Caribbean sea level pressure. Wea. and Forecasting, 13, 740752.
Kossin, J. P., and D. J. Vimont, 2007: A more general framework for understanding Atlantic hurricane
variability and trends. Bull. Amer. Meteor. Soc., 88, 1767-1781.
Landsea, C. W., 1991: West African monsoonal rainfall and intense hurricane associations. Dept. of
Atmos. Sci. Paper, Colo. State Univ., Ft. Collins, CO, 272 pp.
Landsea, C. W., 1993: A climatology of intense (or major) Atlantic hurricanes. Mon. Wea. Rev., 121,
1703-1713.
Landsea, C. W., 2007: Counting Atlantic tropical cyclones back to 1900. EOS, 88, 197, 202.
Landsea, C. W. and W. M. Gray, 1992: The strong association between Western Sahel monsoon rainfall
and intense Atlantic hurricanes. J. Climate, 5, 435-453.
Landsea, C. W., W. M. Gray, P. W. Mielke, Jr., and K. J. Berry, 1992: Long-term variations of Western
Sahelian monsoon rainfall and intense U.S. landfalling hurricanes. J. Climate, 5, 1528-1534.
55
Landsea, C. W., W. M. Gray, K. J. Berry and P. W. Mielke, Jr., 1996: June to September rainfall in the
African Sahel: A seasonal forecast for 1996. 4 pp.
Landsea, C. W., N. Nicholls, W.M. Gray, and L.A. Avila, 1996: Downward trends in the frequency of
intense Atlantic hurricanes during the past five decades. Geo. Res. Letters, 23, 1697-1700.
Landsea, C. W., R. A. Pielke, Jr., A. M. Mestas-Nunez, and J. A. Knaff, 1999: Atlantic basin hurricanes:
Indices of climatic changes. Climatic Changes, 42, 89-129.
Landsea, C.W. et al., 2005: Atlantic hurricane database re-analysis project. Available online at
http://www.aoml.noaa.gov/hrd/data_sub/re_anal.html
Mielke, P. W., K. J. Berry, C. W. Landsea and W. M. Gray, 1996: Artificial skill and validation in
meteorological forecasting. Wea. Forecasting, 11, 153-169.
Mielke, P. W., K. J. Berry, C. W. Landsea and W. M. Gray, 1997: A single sample estimate of shrinkage
in meteorological forecasting. Wea. Forecasting, 12, 847-858.
Pielke, Jr. R. A., and C. W. Landsea, 1998: Normalized Atlantic hurricane damage, 1925-1995. Wea.
Forecasting, 13, 621-631.
Rasmusson, E. M. and T. H. Carpenter, 1982: Variations in tropical sea-surface temperature and surface
wind fields associated with the Southern Oscillation/El Niño. Mon. Wea. Rev., 110, 354-384.
Seseske, S. A., 2004: Forecasting summer/fall El Niño-Southern Oscillation events at 6-11 month lead
times. Dept. of Atmos. Sci. Paper No. 749, Colo. State Univ., Ft. Collins, CO, 104 pp.
Vimont, D. J., and J. P. Kossin, 2007: The Atlantic meridional mode and hurricane activity. Geophys. Res.
Lett., 34, L07709, doi:10.1029/2007GL029683.
Wheeler, M. C., and H. H. Hendon, 2004: An all-season real-time multivariate MJO index: Development
of an index for monitoring and prediction. Mon. Wea. Rev., 132, 1917-1932.
Wolter, K, and M. S. Timlin, 1998: Measuring the strength of ENSO events – how does 1997/98 rank?
Weather, 53, 315-324.
56
14
Verification of Previous Forecasts
Table 18: Summary verification of the authors’ six previous years of seasonal forecasts for Atlantic TC
activity between 2004-2009.
2004
Named Storms
Named Storm Days
Hurricanes
Hurricane Days
Major Hurricanes
Major Hurricane Days
Net Tropical Cyclone Activity
5 Dec. 2003
13
55
7
30
3
6
125
Update
2 April
14
60
8
35
3
8
145
Update
28 May
14
60
8
35
3
8
145
Update
6 August
13
55
7
30
3
6
125
Obs.
15
93
9
45.50
6
22.25
232
2005
Named Storms
Named Storm Days
Hurricanes
Hurricane Days
Major Hurricanes
Major Hurricane Days
Net Tropical Cyclone Activity
3 Dec. 2004
11
55
6
25
3
6
115
Update
1 April
13
65
7
35
3
7
135
Update
31 May
15
75
8
45
4
11
170
Update
5 August
20
95
10
55
6
18
235
Obs.
28
131.50
15
49.75
7
17.75
279
2006
Named Storms
Named Storm Days
Hurricanes
Hurricane Days
Major Hurricanes
Major Hurricane Days
Net Tropical Cyclone Activity
6 Dec. 2005
17
85
9
45
5
13
195
Update
4 April
17
85
9
45
5
13
195
Update
31 May
17
85
9
45
5
13
195
Update
3 August
15
75
7
35
3
8
140
Obs.
10
52.75
5
21.25
2
2
85
2007
Named Storms
Named Storm Days
Hurricanes
Hurricane Days
Major Hurricanes
Major Hurricane Days
Net Tropical Cyclone Activity
8 Dec. 2006
14
70
7
35
3
8
140
Update
3 April
17
85
9
40
5
11
185
Update
31 May
17
85
9
40
5
11
185
Update
3 August
15
75
8
35
4
10
160
Obs.
15
37.75
6
12.25
2
6
99
2008
Named Storms
Named Storm Days
Hurricanes
Hurricane Days
Major Hurricanes
Major Hurricane Days
Accumulated Cyclone Energy
Net Tropical Cyclone Activity
7 Dec. 2007
13
60
7
30
3
6
115
125
Update
9 April
15
80
8
40
4
9
150
160
Update
3 June
15
80
8
40
4
9
150
160
Update
5 August
17
90
9
45
5
11
175
190
Obs.
16
88.25
8
30.50
5
7.50
146
162
2009
Named Storms
Named Storm Days
Hurricanes
Hurricane Days
Major Hurricanes
Major Hurricane Days
Accumulated Cyclone Energy
Net Tropical Cyclone Activity
10 Dec. 2008
14
70
7
30
3
7
125
135
Update
9 April
12
55
6
25
2
5
100
105
Update
2 June
11
50
5
20
2
4
85
90
Update
4 August
10
45
4
18
2
4
80
85
Obs.
9
30
3
12
2
3.50
53
69
57